RefereraExporteraLink to record
Permanent link

Direct link
Referera
Referensformat
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annet format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annet språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
Increasing Rule Extraction Accuracy by Post-processing GP Trees
Högskolan i Borås, Institutionen Handels- och IT-högskolan.ORCID-id: 0000-0003-0412-6199
Högskolan i Borås, Institutionen Handels- och IT-högskolan.
Högskolan i Borås, Institutionen Handels- och IT-högskolan.ORCID-id: 0000-0003-0274-9026
Högskolan i Borås, Institutionen Handels- och IT-högskolan.
2008 (engelsk)Inngår i: Proceedings of the Congress on Evolutionary Computation, IEEE, 2008, s. 3010-3015Konferansepaper, Publicerat paper (Fagfellevurdert)
Abstract [en]

Genetic programming (GP), is a very general and efficient technique, often capable of outperforming more specialized techniques on a variety of tasks. In this paper, we suggest a straightforward novel algorithm for post-processing of GP classification trees. The algorithm iteratively, one node at a time, searches for possible modifications that would result in higher accuracy. More specifically, the algorithm for each split evaluates every possible constant value and chooses the best. With this design, the post-processing algorithm can only increase training accuracy, never decrease it. In this study, we apply the suggested algorithm to GP trees, extracted from neural network ensembles. Experimentation, using 22 UCI datasets, shows that the post-processing results in higher test set accuracies on a large majority of datasets. As a matter of fact, for two setups of three evaluated, the increase in accuracy is statistically significant.

sted, utgiver, år, opplag, sider
IEEE, 2008. s. 3010-3015
Emneord [en]
genetic programming, rule extraction, Computer Science, Machine Learning, Data Mining
Emneord [sv]
data mining
HSV kategori
Identifikatorer
URN: urn:nbn:se:hj:diva-45814Lokal ID: 0;0;miljJAILISBN: 978-1-4244-1823-7 (tryckt)OAI: oai:DiVA.org:hj-45814DiVA, id: diva2:1348931
Konferanse
CEC 2008, Hong Kong, June 1-6, 2008
Merknad

Sponsorship:

This work was supported by the Information Fusion Research Program (University of Skövde, Sweden) in partnership with the Swedish Knowledge Foundation under grant 2003/0104 (URL: http://www.infofusion.se).

Tilgjengelig fra: 2019-09-06 Laget: 2019-09-06 Sist oppdatert: 2019-09-06bibliografisk kontrollert

Open Access i DiVA

fulltext(155 kB)25 nedlastinger
Filinformasjon
Fil FULLTEXT01.pdfFilstørrelse 155 kBChecksum SHA-512
f3248da0eb4b934a15ce0c29be865e90b34985095f81c0ee8d70f26bc5a89964a24b678b22b8e476115225fc46be9477bb27d59ae005409f792bf948075c00c6
Type fulltextMimetype application/pdf

Personposter BETA

Johansson, UlfKönig, RikardLöfström, Tuve

Søk i DiVA

Av forfatter/redaktør
Johansson, UlfKönig, RikardLöfström, Tuve

Søk utenfor DiVA

GoogleGoogle Scholar
Totalt: 25 nedlastinger
Antall nedlastinger er summen av alle nedlastinger av alle fulltekster. Det kan for eksempel være tidligere versjoner som er ikke lenger tilgjengelige

isbn
urn-nbn

Altmetric

isbn
urn-nbn
Totalt: 62 treff
RefereraExporteraLink to record
Permanent link

Direct link
Referera
Referensformat
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annet format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annet språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf